Introduction
Social Security, the largest social safety-net program in the United States,
is funded by a 12.4% flat tax on wages capped at $160,200.
Because earnings above this wage cap are exempt from tax, Social Security tax collection is regressive, with low-wage taxpayers paying a higher average tax rate on their wages relative to taxpayers with wages exceeding this threshold.
This leads to peculiar outcomes: As salaries exceed the wage cap, the Social Security tax rate decreases and approaches zero.
To address this normative shortcoming, this Piece proposes eliminating the wage cap and using the resulting additional revenue to fund a zero-rate Social Security tax bracket analogous to the standard deduction of the federal income tax. IRS data show that eliminating the wage cap could fund an exemption of at least $10,000, thereby making Social Security tax collection progressive across all wage earners. By altering only the distribution of the tax burden but not the total revenue collected or the Social Security benefits, these changes would also be straightforward to implement. This benefit-neutral and revenue-neutral change to Social Security would be especially valuable for the millions of impoverished low-wage earners for whom the additional funds would provide much-needed support and for certain racial groups whose decreased life expectancy reduces their total Social Security benefits.
Part I of this Piece provides an overview of Social Security. Part II discusses Social Security’s regressivity and explains why Social Security tax collection and benefit provision should both be progressive. Part III discusses the proposal’s particulars, namely, eliminating the wage cap for Social Security taxes and using the revenue generated to create a zero-rate Social Security tax bracket.
I. Social Security: An Overview
The Social Security program was launched in 1935 in response to the hardships caused by the Great Depression, with its best-known benefit providing payments to retirees aged sixty-five and older.
Although the program has evolved since its inception, several original features remain, including compulsory participation, calculation of benefits as a function of payments, and financing of benefits from payroll taxes levied on employees and employers.
Today, Social Security provides monthly benefits to about 66 million people, making annual payments in excess of $1.2 trillion.
Taxes on wages fund Social Security benefits.
Employers and employees together pay 12.4% of the first $160,200
in salary, primarily to two separate trust funds: Old-Age and Survivors Insurance (OASI) and Disability Insurance (DI).
OASI benefits retired workers and their dependents; DI benefits disabled workers and their dependents.
Together, these two trust funds and their associated benefits comprise what is colloquially known as Social Security.
Although they are technically separate entities, they are commonly presented as one fund for actuarial purposes: the OASDI.
Social Security differs from most other federal spending programs in that payments for beneficiaries come from the OASDI trust fund, not the federal government’s general funds.
Congress is obligated to make payments to beneficiaries and is precluded from using OASDI funds for any other purpose.
Social Security is thus protected from the congressional scavenging to which general funds might be vulnerable. Social Security taxes paid by current wage earners go directly into the trust fund, and later to the current beneficiaries.
Benefit payments to retirees comprise the lion’s share of OASDI payouts.
These payouts are calculated by reference to average indexed monthly earnings (AIME), an inflation-adjusted measure of a worker’s average income during the thirty-five years of their highest earnings.
The income credited for any given year is capped at the wage cap in place for that year.
The SSA’s method of calculating a retiree’s monthly Social Security benefit weights higher AIMEs proportionately less than lower AIMEs.
Even though a higher AIME always results in a higher total Social Security benefit, the Social Security benefit as a percentage of AIME decreases as AIME increases. Thus, Social Security benefits (in contrast to Social Security taxes collected) are progressive, with monthly benefits as a percentage of AIME decreasing as wages increase.
Although Social Security benefits are progressive (with respect to AIME) for beneficiaries, this is not sufficient to ensure progressivity of the Social Security program in the aggregate. The following Part describes how Social Security is markedly more regressive if assessed by income and explains how progressivity of benefits is not sufficient to counteract the regressivity of Social Security tax collection.
I. The Regressivity of Social Security Taxes
A. Social Security Tax Collection
Taxpayers with wages above the Social Security wage cap pay a lower average Social Security tax rate than taxpayers with earnings below the wage cap.
Because wages above the cap are exempt from tax (that is, there is a zero marginal Social Security tax rate imposed on wages above the cap), the average Social Security tax rate imposed on these earners will necessarily be less than the average rate imposed on earners with wages below the cap. A lawyer earning a salary of $300,000 is taxed at approximately half the rate of a nurse earning $100,000,
while a college football coach earning $10 million pays a social security tax rate of approximately 0%.
Even though the lawyer and the football coach have the same Social Security tax liability as measured in total dollars—$19,865—this fixed liability represents a decreasing percentage of wages as wages increase.
Progressivity assessments of Social Security often use wages as both the taxable base and the distributional yardstick by which the tax burden is compared.
Because the tax burden (expressed as a percent of wages) decreases as wages increase beyond the wage cap, Social Security taxes can be considered regressive. This analysis compares taxpayers using wages, but if regressivity were measured with reference to income, the distributional shortcomings of Social Security tax collection become even more pronounced and Social Security taxes would be significantly more regressive. Because high-wage earners also tend to have other sources of income, such as capital gains and other returns from investment, the percentage of income paid in Social Security taxes for these high-wage earners is generally much lower than the percentage of wages.
For low-wage earners, wages and total income are essentially identical.
Thus, to the extent income (as opposed to wages) is a better measure of a taxpayer’s ability to pay and therefore a better metric by which to assess progressivity, the current Social Security tax rate structure is even more regressive than generally considered. If the lawyer from the preceding example has $100,000 in capital gains, for instance, the Social Security tax rate as a percent of income drops from 6.6% to 5.0%.
B. Social Security Benefits
Social Security’s benefit formula, as described in Part I, credits the first dollars of a worker’s earnings at a much higher percentage than their last dollar and is therefore progressive with respect to AIME. In other words, a retiree with double the credited lifetime earnings will not necessarily receive double the Social Security benefit. For example, using numbers from 2022, a retiree whose AIME is $1,000 receives a monthly benefit of $900;
a retiree with an AIME of $2,000 receives a monthly benefit of only $1,234.
Although AIME has doubled, retirement benefits have only increased by 37%.
The progressive nature of Social Security benefits mitigates the regressivity of Social Security tax collection.
But there is not scholarly consensus on whether the progressivity of the payout structure is sufficient to counteract the regressivity of the rest of the program.
For any individual retiree, there is no guarantee that the retiree will obtain the benefits for which they have contributed. A worker who dies the day after retiring may have paid into the Social Security program over the course of their entire working life but will receive zero benefits. Because low-wage earners generally have shorter lifespans, the progressivity of the Social Security payout structure (which is a lifetime benefit) is muted by the shorter payout schedule for low-wage earners.
Discrepancies in life expectancy also disproportionately affect certain racial minorities. According to a recent report by the NIH, the life expectancy for Black Americans and American Indians is approximately four years and six years lower, respectively, than the seventy-nine-year life expectancy for white Americans.
Given that full Social Security retirement benefits are available only at age sixty-seven, Black and American Indian retirees with full benefits would receive only seventy percent and fifty-one percent of the retirement benefits of white retirees, respectively.
Changing the distribution of Social Security tax collection would not alter these discrepancies in benefit payments. But since these racial groups generally earn less than white workers,
decreasing the regressivity of Social Security tax collection would reduce the price that these racial groups pay for these truncated benefits.
Focusing on the progressivity of Social Security benefits thus masks the normative shortcomings of the program as a whole. Progressivity should be a feature not only of payouts but also of Social Security tax collection. The following Part discusses how removing the existing wage cap and creating a zero-rate Social Security tax bracket would accomplish that goal.
III. Removing the Social Security Wage Cap to Fund a Zero-Rate Bracket
This Piece proposes removing the wage cap on Social Security taxes and using the additional revenue to create a zero-rate bracket for wages earned. Both wage earners and employers would pay Social Security taxes on all wages paid, not just wages below the wage cap. This exemption would function similarly to how the existing standard deduction currently operates for the federal income tax. Using IRS data, this Piece calculates that the resulting additional revenue raised would fund a zero-rate bracket excluding at least the first $10,000 of wages from Social Security taxes for all earners.
Taxpayers with more than $10,000 of wages would receive an approximately $1,200 benefit.
Taxpayers with less than $10,000 of wages would pay no Social Security taxes.
Eliminating the wage cap and creating a zero-rate bracket transforms Social Security tax collection from regressive to progressive.
By giving all workers some amount of wages that are exempt from tax, the average Social Security tax rate will start at 0% and approach 12.4%, rather than start at 12.4% and approach 0%.
Under the existing system, even if the progressivity of Social Security benefit payments mitigates the regressivity of collection, the uncertainty over whether any specific taxpayer will actually receive the benefits for which they have paid means that progressivity is not assured for all taxpayers. As a normative matter, this Piece asserts that Social Security tax collection and benefit provision should both be progressive, rather than just the latter.
Notably, the Social Security benefits that each worker obtains would not change. By explicitly connecting the size of the zero-rate bracket to the revenue raised from eliminating the wage cap, there is no risk that the proposal will jeopardize existing payment obligations to retirees. Wage earners would continue to receive credit based on their AIME, which would continue to be limited by the wage cap (adjusted for inflation) in place for that year. Rather than changing Social Security benefits, the proposal changes the distribution of the Social Security tax burden by shifting payments from low-wage earners to high-wage earners. By keeping elements of Social Security other than tax collection constant, this proposal would be administratively straightforward to implement.
A. Implementation
To create the zero-rate bracket, the IRS would use the additional Social Security taxes collected over the course of the taxable year to pay for an exemption for taxpayers’ wages from that tax year. Employers currently remit employment taxes to the IRS every quarter, with all employment taxes for the previous year generally filed by January 31 following the close of that taxable year.
Thus, the revenue available to create the zero-rate bracket for that taxable year will be known prior to processing taxpayers’ income tax returns for that year.
The income tax filings currently submitted by taxpayers already include all the information needed to implement the zero-rate bracket. Taxpayers’ wages are reported on line 1a of Form 1040 and are also documented on the Form W-2 sent by each employer to the IRS.
Taxpayers would, upon filing their income tax return, obtain a refund for the Social Security taxes previously levied on wages within the exemption amount. Assuming an exemption amount of $10,000, workers with wages of $50,000 and $5,000 would receive Social Security tax refunds of approximately $1,200 and $600, respectively.
Because the zero-rate bracket is intended to be revenue neutral, there will be a time lag between collecting the tax revenues needed to fund the zero-rate bracket and determining the zero-rate bracket’s exact size. That is, the size of the zero-rate bracket will be known only after the additional revenue generated from lifting the wage cap is known. The IRS, using its internal models, could reasonably approximate the zero-rate bracket’s size and release preliminary estimates as desired. Although precision would decline, giving workers real-time increases to their paychecks might be better than requiring workers to file before obtaining the benefit. Alternatively, the zero-rate bracket could start the year following the elimination of the cap: By using tax revenue from year one to fund the zero-rate bracket in year two, the IRS could release the exact size of the zero-rate bracket well before the end of the taxable year.
Eliminating the wage cap would improve certain aspects of tax administration. The current system can result in an overcollection of Social Security taxes for an employee with multiple employers if the employee’s aggregate wages exceed the wage cap.
Because the proposal subjects all wages to Social Security tax and confers the benefits of the zero-rate bracket only upon filing, each employer’s Social Security tax obligations would be independent of the obligations of other employers. Additionally, employers would no longer need to adjust withholding rates during the tax year at the moment their employees’ wages exceed the cap.
B. Estimates of Revenue Generated and Size of the Exemption
Using publicly available IRS data, it is possible to estimate the revenue generated from eliminating the wage cap and approximate the size of the wage exemption that the additional revenue could fund.
In 2018, for example, $885 billion of Social Security taxes were collected, which represented taxation of 83% of total wages.
Eliminating the wage cap would thus result in approximately $181 billion with which to create a zero-rate bracket.
That same year there were approximately 145 million wage earners,
meaning that eliminating the wage cap in 2018 could have funded a zero-rate bracket of at least $10,000 across all wage earners,
with taxpayers earning wages of $10,000 or more saving approximately $1,200.
IRS data for the years following 2018 are limited, but by making reasonable assumptions, the size of the zero-rate bracket can be estimated for additional years. Table 1 provides estimates of the zero-rate bracket for 2019, 2021, and 2022—years for which the data are less complete.
Table 1. Estimated Zero-Rate Wage Tax Bracket Funded by Eliminating the Social Security Wage Cap
| Year |
Taxable Ratio
|
Total Social Security Tax Contributions (Billions)
|
Additional Social Security Revenue if No Wage Cap (Billions)
|
Number of Wage Earners (Millions)
|
Social Security Taxes Saved per Wage Earner (Dollars)
|
Zero-Rate Bracket Size (Dollars)
|
| 2015 |
82.6% |
794.9 |
167.4 |
144.1 |
1,162 |
9,371 |
| 2016 |
82.7% |
836.2 |
174.9 |
144.2 |
1,213 |
9,783 |
| 2017 |
83.2% |
873.6 |
176.4 |
145.8 |
1,210 |
9,757 |
| 2018 |
83.0% |
885.1 |
181.3 |
144.6 |
1,254 |
10,111 |
| 2019 |
83.0% |
944.5 |
193.5 |
148.2 |
1,305 |
10,527 |
| 2020 |
82.4% |
1,001.3 |
213.9 |
157.5 |
1,358 |
10,951 |
| 2021 |
81.0% |
980.6 |
230.0 |
153.6 |
1,497 |
12,077 |
| 2022 |
82.3% |
1,106.6 |
238.0 |
160.6 |
1,482 |
11,951 |
The years following the onset of the COVID-19 pandemic resulted in a significant decrease in the taxable ratio, with the percent of wages below the wage cap falling from 82.2% to 81% between 2020 and 2021. Because the pandemic increased wage inequality, more wages were earned by taxpayers exceeding the wage cap relative to taxpayers below the cap. If the wage cap would have been eliminated during those years (as this Piece proposes), it would have resulted in proportionately more revenue since a higher percentage of wages would have been taxed. Since this revenue would have been used entirely to fund a zero-rate bracket, the result would have been a greater redistribution of money from high-wage earners to low-wage earners as pre-tax wage inequality increased, thereby mitigating (post-tax) the effects of this inequality.
C. Concerns
1. Efficiency. — To the extent that eliminating the wage cap results in significant behavioral changes, these costs might outweigh the normative improvements of the zero-rate bracket. Eliminating the wage cap, however, is not likely to result in significant inefficiencies (i.e., in workers trading leisure for employment). First, increased taxes only induce behavioral changes to the extent that they are salient to taxpayers—that is, to the extent that taxpayers are aware of the increase.
Social Security taxes (1) are withheld and remitted by employers; (2) only decrease after exceeding approximately $160,000; and (3) are relatively low-rate.
Social Security taxes are thus considered to be of low salience.
Second, efficiency costs arising from changed behavior would arise only if employees were able to limit their wage earnings in response to the additional tax. Even though employees might desire to change their behavior, the realities of lucrative employment practically preclude it. The overwhelming majority of employees earning above $160,200 are salaried rather than hourly and would thus be unable to adjust their wages in response to the increased tax.
It is difficult, for instance, for a football coach to work fewer hours by forgoing a portion of their salary, even if they wished to do so.
Subjecting all wages to Social Security taxes would also likely have a small effect on labor decisions at the extensive margin, that is, on workers’ decisions to enter the labor force. Unlike the income tax’s notable effect on married spouses’ decisions to obtain paid labor (which subjects the spouse to the marginal tax rate of their partner), eliminating the wage cap would (1) only affect earners making over $160,200, and (2) impose a relatively small tax burden. That is, a worker with the potential to earn $200,000 in wages would likely not forgo those earnings because of the $4,900 in additional Social Security taxes owed.
2. Conflicts With Other Changes to Social Security. — The Social Security program is under severe budgetary strain. Under current revenue and benefit predictions, the OASDI trust fund will be exhausted in 2033.
Although the federal government is statutorily obligated to pay benefits to retirees, once the OASDI trust fund is exhausted, these funds must come from other sources, such as the general fund.
If Social Security is intended to be a self-funded program in which benefits to retirees are funded solely by Social Security taxes (plus interest earned and taxes paid on Social Security benefits), significant structural changes to Social Security must be made.
This Piece’s proposal is not intended to substitute for other, deeper Social Security reforms. Rather, it addresses a specific shortcoming of the existing Social Security program—its regressivity—in an administratively straightforward manner without altering other elements of the program. The proposal’s key features, elimination of the wage cap in conjunction with the creation of a zero-rate bracket, do not preclude other fundamental changes to Social Security that Congress deems necessary. Funding the OASDI trust fund with additional taxes on investment income and business income, for instance (as recently proposed by Senators Bernie Sanders and Elizabeth Warren), can be done in conjunction with this Piece’s proposal.
3. Unintended Beneficiaries. — This Piece’s proposal is premised on the assumption that redistributing Social Security tax burdens from low-wage earners to high-wage earners is, as a normative matter, preferred. Assuming the declining utility of dollars for higher-earning taxpayers, redistributing a dollar from a taxpayer with wages above $160,200 to a taxpayer earning less than that should result in an increase of overall welfare. For low-wage taxpayers, the savings from the zero-rate bracket would function similarly to the Earned Income Tax Credit, which lifts millions of the working poor out of poverty.
By using wages (rather than income or wealth) as the distributional base, however, there is the potential for certain disfavored groups to benefit from the newly formed zero bracket. Part-time workers, including students and the elderly, would also have their first $10,000 of wages exempted from Social Security taxes. But to the extent this is problematic (for example, because low wages don’t accurately reflect their status as the working poor because they are rich retirees or students with significant future earning potential), Congress can easily tailor the zero-rate bracket for groups meeting the desired characteristics. Because the zero-rate bracket will be implemented via income tax returns, Congress could use attributes contained within income tax filing to limit eligibility. Full-time students, taxpayers claimed as dependents, or taxpayers below or above a certain age, for example, could all be precluded from using the exemption if Congress so desired.
Conclusion
By eliminating the wage cap on Social Security taxes and using the additional revenue to subsidize a zero-rate wage bracket, Social Security tax collection can be made progressive without altering the benefits that current retirees receive and future beneficiaries expect. By so doing, roughly ninety-four percent of U.S. taxpayers will pay less in Social Security taxes with no reduction in benefits.
Millions of low-wage taxpayers will receive a credit for Social Security taxes paid, thus expanding Social Security’s promise to provide a social safety net—not just for retirees but for workers in the present as well.